Hydrology Infiltration

4. INFILTRATION AND ITS ESTIMATION

4.1 Infiltration Process

Infiltration is as the entry or the passage of water into the soil through soil
surface. It is a major loss of precipitation affecting runoff of a basin. This term
should be properly understood and quantified. Losses like interception,
depression storage and evaporation during precipitation are small, which cannot
change the runoff of a basin significantly during major floods, but infiltration is a
major process continuously affecting the magnitude, timing and distribution of
surface runoff at any measured outlet of a basin. It is responsible for the growth
and nourishment of life on earth. Infiltration process is initiated by creation of
hydrogen bond between soil particles and the water. The adhesive force of
attraction between soil and water, the surface tension, capillarity and
gravitational forces help to force more water between the pores of soil particles
as more water is added to the system due to rain.

Infiltration capacity is the maximum rate at which a given soil can absorb water
under a given set of conditions at a given time. At any instant the actual
infiltration f t can be equal to infiltration capacity fo only when the rainfall intensity
is greater than fo, otherwise actual infiltration will be equal to the rate of rainfall.
This can be observed during low intensity rainfall when there is no surface runoff
produced due to precipitation. Once water enters into the soil, the process
transmission of water within the soil known as percolation takes place, thus
moving the water from near the surface to down below, charging the ground
water reservoir. Infiltration and percolation are directly interrelated. When
percolation stops, infiltration also stops. During any storm, infiltration is the
maximum at the beginning of the storm, decays exponentially and attains a
constant value J, as the storm progresses. The effect of infiltration is to
(i) reduce flood magnitude
(ii) delay the time of arrival of water to the channel ..
(iii) reduce the soil erosion

AMU 1 HWRE
Hydrology Infiltration

(iv) recharge to the ground water reservoir

(v) fill the soil pores with water to its field-capacity, which subsequently
supply water to the plants
(vi) avail the ground water during the non-rain periods in the channels
(vii) help to supply water to plants

4.2 Factors Affecting Infiltration

Factors affecting infiltration depend on both meteorological and soil medium

characteristics. These are:
(a) Surface Entry
If a soil surface is bare, the impact of raindrops causes in washing of finer
particles and clogs the surface. This retards infiltration. An area covered by grass
and other bushy plants has better infiltration capacity than a barren land.
(b) Percolation
For infiltration to continue, water that has entered the soil must be transmitted
down by the force of gravity and capillary actions. When percolation rate (Pr) is
slow, the infiltration rate is bounded by the rate of percolation. This depends on
the factors like type of soil, its composition, permeability, porosity, stratification,
presence of organic matter, and presence of salts. A graph between infiltration
capacity vs. grain size of the medium is given in Fig. 4.1.

Fig 4.1 Infiltration Capacity vs Medium Grain Size

(c) Antecedent Moisture Condition
Infiltration depends on the presence of moisture in the soil. For the second storm
AMU 2 HWRE
Hydrology Infiltration

in succession, the soil will have lesser rate of infiltration than the first maiden
storm of the season. Except sandy soil most other soils have swelling
ingredients, which swells in presence of water and reduce infiltration rate to the
extent of their presence.

(d) Climate Conditions

Temperature affects the viscosity of water. Variation of infiltration with soil tem-
perature is shown in Fig. 4.2. Flow of water within the body of the soil is laminar;
the flow being directly related to viscosity. In summer, therefore, less viscous
water causes more infiltration than in winter. In sub-zero temperature, water
present in soil pores gets crystallized, thus blocking the passage. Other climatic
factors may not influence infiltration rate to the extent, temperature does, and
therefore, temperature can be considered as the only vibrant climatic factor
affecting infiltration.

Fig 4.2 Variation of Soil Temperature and Infiltration with Time

(e) Rainfall Intensity and Duration

During heavy rainfall, the top soil is affected by mechanical compaction and by in
wash of finer materials. This leads to faster decrease in the rate of infiltration
than with low intensities of rainfall. Duration of rain affects to the extent that when

AMU 3 HWRE
Hydrology Infiltration

the same quantity of rain falls in n number of isolated storms instead of a

continuous one, the infiltration will be higher in the former case.

(f) Human Activities

When crops are grown or grass covers a barren land, the rate of infiltration is
increased. On the other hand construction of roads, houses, Overgrazing of
pastures and playgrounds reduce infiltration capacity of an area considerably.

(g) Depletion of Ground Water Table

Position of ground water table should not be very close to the surface for
infiltration to continue. The quantity of infiltrated water entering into the soil
should be drained out fully from the top soil zone so that there is some space
available for the infiltrated water to store during the next rain.

(h) Quality of Water

Water containing silt, salts and other impurities affect the infiltration to the extent
they are present. Salts present affect the viscosity of water and may also react
chemically with soil to form complexes which obstruct the porosity of soil, thereby
affecting infiltration. Silts clog the pore spaces retarding infiltration rate
considerably.

(i) Vegetation
Soil covered with vegetation has greater infiltration than barren land. Because of
growth and decay of roots and bacterial activities, dense natural forests provide
good infiltration than sparsely planted crops.

(j) Grain Size of Soil Particles

Other factors remaining the same, infiltration rate is directly proportional to the
grain size diameter. When swelling minerals like illite and montmorillonite are
present in soils, the infiltration rate reduces.

(k) Catchment Parameters

AMU 4 HWRE
Hydrology Infiltration

A correlation between the drainage density (length of all streams of a basin

divided by its catchment area) and infiltration can be established for various
basins. Such basins exhibit negative correlation. When the drainage density
increases, infiltration capacity decreases. A similar type of correlation can be
obtained between the runoff rate (m3/ sq. km) and infiltration capacity (cm/h). A
relation between the sediment yield rate in ha.m/sq. km and infiltration capacity in
cm/h also exhibit an inverse proportionality.

Three distinct approaches to determine infiltration capacity of a basin are

available.
(i) Experiment by a single or double cylindrical infiltrometer
(ii) Rainfall simulator
(iii) Rainfall-runoff analysis.

4.3 Infiltration Measurements

4.3.1 Field Measurement Using Infiltrometers

Two types of infiltrometers used are: single cylindrical and concentric double
cylindrical.

Single Tube lnfiltrometer

It consists of a hallow metal cylinder 30 cm in diameter and 60 cm long driven
into the ground such that 10 cm of it projects above ground level. Water is
poured at the top such that a head of 7 cm within the infiltrometer is maintained
above ground level. A graduated jar or burette is used to add water ( Fig. 4.3), to
give directly the volume of water added over time. A plot of time in abscissa
against rate of water added in mm/h gives an infiltration capacity curve for the
area. The setup resembles to the flooding type of irrigation situation in the field
which can be a possible representation of real local conditions. Sufficient
precautions should be taken to drive the cylinder into the ground with minimum
disturbance to the soil structure. In a single infiltrometer, the major criticism is

AMU 5 HWRE
Hydrology Infiltration

that water spreads out immediately beyond the bottom of the cylinder which does
not represent a true infiltration condition of the field.

Fig 4.3 A Single Tube Infiltrometer

Double Tube lnfiltrometer

To overcome the objections of a single ring infiltrometer a set of two concentric
hollow cylinders of same length are used (Fig. 4.4). Water is added to both the
rings to maintain the same height. Reading of the burette for the inner cylinder is
taken as infiltration capacity of the soil. The outer cylinder is maintained to
prevent spreading of water from the inner one.

AMU 6 HWRE
Hydrology Infiltration

Fig 4.4 A Double Tube Infiltrometer

The important disadvantages still prevalent in these types of infiltrometers are:

(i) The size effect: Larger diameter infiltrometers give more accurate and
always lesser value of infiltration than smaller diameter type.
(ii) Boundary effect.
(iii) Disturbance of original soil due to driving of the rings.
In the market a single ring portable infiltrometer is available which gives
infiltration results quite acceptable for all practical purposes.

Rainfall Simulator
Rainfall simulators are used to overcome the difficulties of infiltrometers. A
rainfall simulator consists of a sprinkler with nozzles capable of producing
artificial rain of various intensities, drop sizes and durations. A field plot of 1.8 m
in width and 3.65 m or more length is selected on which the nozzles spray water
at height of 2 m or more to the field. The terminal velocity of rainfall is assumed
to be close to raindrops though a height of 5-6 m may be the ideal situation.
Rainfall intensities of 44.5 mm/h or multiple of it are generally created under such
conditions. Arrangement is made to collect the runoff from the plot which can be
measured (SRd). Prior to the experiment a run is made covering the plot with a

AMU 7 HWRE
Hydrology Infiltration

polyethylene sheet to know the average rate of rainfall (Pd). Test run starts after
removal of the polythene cover from the field and continues till a steady state of
runoff from the plot is obtained. The following water budget equation is used to
estimate infiltration rate from the experiment.

Fd = Pd - SRd - Sol (4.1)

where Fd is the depth of infiltrated water in mm (to be computed), Pd the

simulated rainfall depth in mm (measured), SRd the surface runoff depth in mm
(measured), Sol the depression storages, surface detention, abstraction and other
losses in mm. When steady state is reached, the analytical run is carried out, the
volume Sol being no more effective, the constant rate of infiltration is calculated
from the relation Fd = (Pd - SRd). When rainfall stops, runoff from the plot
continues and the value of depression storage plus surface detention can be
measured from the recession between stop of rainfall and the last drop of water
owing out of the plot as runoff. Abstraction losses (if any) can also be measured
from the plot at the initial stage. Normally rainfall simulator gives less rate of
infiltration than the flooding type infiltrometers.

4.3.2 Rainfall-Runoff Analysis

A plot between runoff from a watershed vs. time is known as Hydrograph.

Analysis of a hydrograph to determine infiltration is a practical field solution,
which considers the variations of rainfall, land forms, land slope, size of
catchment, soil characteristics, vegetation types, depression storage, surface
detention and all other factors affecting infiltration and as they occur in a field. A
network of rain gauges gives the average storm precipitation over a basin. The
runoff from a basin can be measured at the outlet or at the desired location by
gauging the stream. The d ifference between precipitation and the corresponding
runoff averaged over the basin over time should give the total loss. This should
form the basis for estimation of infiltration. During storms, evapotranspiration is
negligible. Depression storage and interception can be estimated or they can be
AMU 8 HWRE
Hydrology Infiltration

considered as lumpsum and subtracted from the rainfall hyetograph at the

beginning of the storm. The balance gives a true representation of the basin
infiltration losses. To simplify the procedure, some theoretical curves are
proposed, which divide the rainfall histogram blocks between runoff and total
losses.
An ideal infiltration capacity curve proposed by Horton (1933, 39) is

ft = fc + (fo – fc) e-kt (4.2)

Here ft is the infiltration capacity at any time t from the beginning of the storm
mm/h, fc the infiltration rate in mm/h at the final steady stage when the soil profile
becomes fully saturated, fo the maximum initial value when t = 0 in mm/h it the
beginning of the storm, k an empirical constant depending on soil cover complex,
vegetation and other factors and t the time lapse from the onset of the storm.
Values of fc, fo and k are dependent on number of factors like soil characteristics
and climatic conditions. Even for the same catchment the values vary from storm
to storm and from season to season. A typical curve of ft separating the rainfall
intensity histogram between infiltration and surface runoff shown in Fig. 4.5.

Fig 4.5 A plot of Infiltration Capacity Curve

AMU 9 HWRE
Hydrology Infiltration

Depending upon the situation, a number of methods are available to accurately

determine the infiltration curve. For small watersheds (for urban hydrology) the
procedure may be entirely different to the analysis carried out for large
watersheds.
Since we are interested in determining the infiltration rate for storms of large
basins, an average infiltration rate known as infiltration index is assumed. This
type of assumption underestimates infiltration rate during the initial part of the
storm and slightly overestimates towards the end of it. The assumption holds
fair1y good for analysis of high intensity and long duration storms or in a
situation, when the area is fairly saturated before the onset of the storm. Flood
producing, storms follow both the situations.
Difficulties with the theoretical approach to infiltration and practical difficulties to
determine the values of fo and k, led to the use of infiltration indices based on an
empirical approach.

4.4 Infiltration Indices

The following indices are mostly used for computation of infiltration rate from
rainfall runoff data.

4.4.1 Φ-1ndex

The average rate of rainfall above which the rainfall volume equals to runoff
volume is called Φ-index. Procedure to derive Φ-index is described below. A
schematic diagram showing the meaning of the Φ-index is given in Fig. 4.6.

AMU 10 HWRE
Hydrology Infiltration

Fig 4.6 Effective Rainfall Using Φ- Index

1. Storm producing runoff for the basin is identified.

2. All rainfall data in and around the basin are collected for the selected
storm along with data of at least one self recording gauge.
3. Corresponding storm runoff data for the basin are also collected.
4. Plot the runoff hydro graph and separate the base flow by any of the
suitable method to estimate the direct runoff hydrograph. Compute the
depth of direct runoff in mm by dividing the area of the basin by the total
volume of the direct runoff. The concept of hydro graph and the separation
of base flow are discussed in chapter 5.
5. Find average aerial rainfall of the basin by any of the suitable method
described in chapter 2.
6. Compute time distribution percentages of storm rainfall on the basis of the
self recording rain gauge data (or average time distribution if more than
one SRRG data are available) for the basin under consideration. This is
required because the non-recording rain gauges record rainfall for 24 h,
whereas we require rainfall intensities over the basin for shorter periods of
2 to 6 h to prepare rainfall intensity histogram.
7. Distribute the average basin rainfall of step (5) to smaller convenient time
unit of say 2 to 6 h according to the percentage distribution of rainfall of
step (6).

AMU 11 HWRE
Hydrology Infiltration

8. Estimate initial losses due to interception and depression storage in mm

and subtract it from the first initial rainfall intensity histogram blocks.
9. Assume a trial value (depth) of Φ -index in mm/h and compute the total
rainfall excess depth by summing up the excess depth of rainfall beyond
this trial index line in the histogram. Compare it with the direct runoff depth
obtained in step (4).
10. If rainfall excess is more, then increase the assumed value of Φ -index of
step (9) and repeat step (9) till rainfall excess in mm equals to direct runoff
depth in mm of step(4).
11. Repeat the procedure for other storms, if under study, to obtain a general
Φ -index value for the basin.

Central Water Commission suggests a relation of the form

Φ=I–R (4.3)
in which R = a I1.2 (4.4)

I is the rainfall intensity in mm/h on a daily (24 h) basis, R the runoff in mm. For
computation of R in equation (4.4), take I in mm/day and for computation of Φ in
equation (4.3), divide R from equation (4.4) by 24 to convert it to mm/h. Values of
coefficient a vary from 0.17 to 0.50. For sandy soils it may be taken as 0.20,
coastal alluvium 0.25 to 0.30, silt 0.35, red and clayey soil 0.40, black cotton soil
0.45 and hilly soil 0.50.

4.4.2 W-Index

This index is considered as an improvement over Φ -index in the sense that

surface storage and interception losses are considered in its computation. It is
defined as the average rate of infiltration which equals to the rate of precipitation
minus surface runoff and retention during time t and is expressed as

AMU 12 HWRE
Hydrology Infiltration

W = (P - SRo - DR)/t (4.5)

where P is the total depth of storm rainfall in mm, SRo the total depth of surface
runoff in mm, DR the total depth of surface retention (depth of depression storage
plus interception loss) and t the time in hour during which rainfall rate exceeds
infiltration rate. Obviously when DR = 0, during heavy and longer storms, W-index
and Φ -index become the same.

Sometimes Wmin-index is used instead of W-index when the soil condition is very
wet so that the soil infiltration rate is almost constant and infiltration is at the
minimum rate for the basin. Both Wmin-index and W-index vary from storm to
storm. For design flood, which is usually very large, the two values are nearly the
same. For practical use, refinement of infiltration index may not be justified due to
the fact that rainfall producing storms are comparatively large and they occur in
wet seasons. Initial losses for such storms are negligible in comparison to the
depth of precipitation from them. The assumptions in their estimation and the
limitations involved in the measurement of the parameters are subject to small
errors which do not necessitate refining indices using a sophisticated model.

On the basis of experience, a more direct and simpler method is invariably

adopted for field problems. While estimating design floods from Probable
Maximum Precipitation (PMP) or Standard Project Storm (SPS), a constant index
value of 1.00 mm/h is usually assumed for some catchments in absence of any
data. This index is deducted from the rainfall intensities to give rainfall excess
producing direct runoff volume for the design flood. This is because the
catchment is assumed to be fully wet by the time the design storm producing
design flood is assumed to be prevalent over the basin. Details of Φ-index
computation are given in examples in the class and also in the subsequent
chapters.

AMU 13 HWRE
Hydrology Infiltration

4.4.3 Fav-Index

Here, we assume an average infiltration loss throughout the storm. The time
period is so selected that only period of excess rain or continuous supply of
infiltration is considered, as shown in Fig. 4.7. The procedure is the same as
discussed under Φ -index method except that the period of Fav is considered on
experience and judgment, which may be beyond the time of rainfall excess as
shown in the figure. An adjustment for period of no-rain may be assumed. The
procedure may be more practical but requires adjustment of time-intensity
histogram on judgment for period of no-rain, initial loss and extra period for Fav
beyond the time of rainfall excess. A good knowledge on soil of the basin,
catchment and storm characteristics is required for such adjustments to be
made. The figure is self explanatory for computation of Fav -index.

Fig 4.7 A Plot of Fav Curve

Additional Methods of estimation of infiltrations.

4.5 Mass Curve Method

For large water sheds, available rainfall and runoff data can be used effectively to
obtain the average infiltration rate of a basin. An accurate estimate of infiltration

AMU 14 HWRE
Hydrology Infiltration

over the storm period for large catchments may be of academic interests but for
field engineers the method generally followed is based on the following:
1. Rainfall data covering the entire basin for the storm is collected. The storm
must have covered the entire basin under consideration.
2. Average storm precipitation over the basin is calculated by Thiessen
polygon or any other suitable method.
3. Cumulative of the basin average rainfall is computed and plotted on a
graph paper ( as mass curve) against the corresponding time in abscissa.
4. Runoff of the basin measured at its outlet is used. Base flow is separated
by a suitable method. Direct runoff hydrograph is divided into blocks of
time units as considered for rainfall histogram and its equivalent depths
found out by dividing the catchment area of the basin.
5. Cumulative depth of runoff of step 4 is computed and plotted in graph
paper of cumulative average rainfall of step 3.
6. The difference in ordinates for different time blocks of the two cumulative
plots of step 3 and 5 are read out from the graphs. This cumulative
difference is plotted in the same graph paper.
7. The plot of cumulative graph of step 6 will be irregular in shape. Smoothen
the graph. This graph represents the cumulative value of infiltration and
other losses.
8. Find the loss rate at different times from the slope of this curve. For storms
producing floods, other losses can be neglected.
9. Plot infiltration rate at different times obtained in step 8 and plot a smooth
curve between the infiltration loss rate and time, which gives an infiltration
rate curve for the storm.

AMU 15 HWRE